Abstract
The hypoxic constraint on peak oxygen uptake (ṀO2,peak) was characterized in rainbow trout over a range of ambient oxygen tensions with different testing protocols and statistical models. The best-fit model was selected using both statistical criteria (R2 and AIC) and the model's prediction of three anchor points for hypoxic performance: critical PO2 (Pcrit), maximum ṀO2 and a new metric, the minimum PO2 that supports 50% of absolute aerobic scope (PAAS-50). The best-fitting model was curvilinear using five strategically selected PO2 values. This model predicted PAAS-50 as 70 mmHg (coefficient of variation, CV=9%) for rainbow trout. Thus, while a five-point hypoxic performance curve can characterize the limiting effects of hypoxia in fish, as envisaged by Fry over 75 years ago, PAAS-50 is a promising metric to compare hypoxic constraints on performance in a standardized manner both within and across fish species.